Furthermore, we quantify ligand-receptor interactions between T cell subsets and their regards to immune system infiltration utilizing a publicly obtainable individual melanoma dataset. cells. To raised know how all cell types within a tumor cooperate to assist in malignant development, Kumar et al. researched communication between cells via receptor and ligand interactions using single-cell data and computational modeling. Graphical Abstract Launch The tumor microenvironment comprises many cell types, including malignant, stromal, and immune system cells. This mobile intricacy of tumors is certainly additional elevated with the heterogeneity of every cell type, BJE6-106 such as different clones of tumor cells or the various subsets of immune cells (Jimnez-Snchez et al., 2017; McGranahan and Swanton, 2017). These various cell types all communicate via ligand-receptor interactions, where the ligand can either be secreted and bind to the receptor in soluble form or be membrane-bound and require physical proximity of the two interacting cell types (Ramilowski et al., 2015). Furthermore, communication between these different cell types is implicated in mechanisms for tumorigenesis, tumor progression, therapy resistance, immune infiltration, and inflammation (Hanahan and Weinberg, 2011). Given BJE6-106 the importance of ligand-receptor interactions on patient outcome, therapeutics that target cell-cell interactions have become a useful tool in clinical practice. For example, the immune checkpoint inhibitor ipilimumab targets the CD28 or CTLA4 interaction, and both pembrolizumab and nivolumab target the PD1 or PDL1 interaction (Pardoll, 2012). Despite the clear success of these therapeutics in several tumor types, the response rates are limited. For instance, only about 20%C25% of patients respond to immuno-oncology drugs (Dempke et al., 2017; Schumacher et al., 2015). This limited response rate is likely because of the complex network of cell-cell interactions present in a tumor microenvironment, our knowledge of which is still incomplete (Sarkar et al., 2016). To better stratify patients for existing therapies as well as to discover interactions that could be targeted, there is a need to more fully understand the spectrum of cell-cell interactions occurring in tumor microenvironments and how these interactions affect outcome. Single-cell RNA sequencing (scRNA-seq) approaches are increasingly being used to characterize both the abundance and functional state of tumor-associated cell types and have provided unprecedented detail of the heterogeneity of the cellular composition (Lavin et al., 2017; Tirosh et al., 2016; Zheng et al., 2017). However, beyond characterizing the cellular composition of a tumor, it is crucial to understand how the different cellular components interact with one another to give rise to emergent tumor behavior. Although examples of using both bulk and single-cell sequencing data to examine cell-cell communication exist (Camp et al., 2017; Choi et al., 2015; Costa et al., 2018; Puram et al., 2017; Skelly et al., 2018; Zhou et al., 2017), techniques for connecting these features to biological outcomes of interest and understanding how these interactions quantitatively relate to specific phenotypic outcomes of interest are still limited. Here we developed an approach to characterize cell-cell communication mediated by ligand-receptor interactions across all cell types in a microenvironment using scRNA-seq data. After assigning cell types based on the scRNA-seq data using a decision tree classifier, our approach quantifies potential ligand-receptor interactions between all pairs of cell types based on their gene expression profiles. We demonstrated how this approach can assess similarities and differences in cell-cell communication between six syngeneic mouse tumor models. We then extended our approach to quantify ligand-receptor interactions in human metastatic melanoma samples. Importantly, we examined the association of individual cell-cell interactions with pathophysiological characteristics of the tumor microenvironment. This work advances conceptual and methodological approaches for gaining insights from single-cell studies of the tumor microenvironment and has applications for discovering effective therapeutic targets and biomarkers for stratification Zfp622 of patient treatment. RESULTS scRNA-seq of Syngeneic Mouse Tumor Models Syngeneic mouse tumor models are frequently used to investigate novel immune-oncology therapeutics (Sanmamed et al., 2016). However, the different models are still incompletely understood with respect to ligand-receptor interactions. We performed scRNA-seq on BJE6-106 tumors from six treatment-naive syngeneic mouse tumor models (B16-F10 melanoma, EMT6 breast mammary carcinoma, LL2 Lewis lung carcinoma, CT26 colon carcinoma, MC-38 colon carcinoma, and Sa1N fibrosarcoma; two samples per tumor model). Because some models are poorly infiltrated by immune cells, we additionally enriched for CD45-positive cells. Notably, measured cell.